[02:35] attente, have you noticed any oddities with rhythmbox with your unity gtk module? My rhythmbox menus are all disabled [02:40] attente, also... "UBUNTU_MENUPROXY= rhythmbox" doesn't work with your module like it did with the old one [06:29] good morning [07:08] Good morning [07:18] hey pitti === vibhav is now known as Guest98376 [08:13] pitti: oh jibel is on holidays? [08:13] those French slackers… :p [08:23] didrocks: je ne sais pas [08:23] * pitti was just watching sabdfl's keynote on MWC [08:24] pitti: keynote, is it online? [08:24] pitti: I just saw the round table [08:24] didrocks: https://plus.google.com/110222816216901496766/posts/e1EVyAk6sj8 [08:24] didrocks: I think we mean the same [08:24] discussion panel [08:25] yep [08:25] ok, I think the keynote is something else [08:35] hey desktopers [08:35] lut didrocks pitti [08:35] bonjour seb128 [08:35] I watched that video yesterday evening ;-) [08:37] salut seb128 [08:39] yay, chromium finished linking overnight after the umpteenth time. shame i don't have enough disk space left to actually finish building the package! [08:40] morning :) [08:41] good morning :) [08:41] well yes it is good, I'm not working tomorrow, so today is my Friday :) [08:42] anything to worry about on raring upgrade, keep being promoted to do a partial upgrade today === Guest98376 is now known as vibhav [08:43] chrisccoulson, hey! how are you? [08:43] czajkowski, good morning [08:43] seb128, not too bad thanks. how are you? [08:46] chrisccoulson, I'm good thanks [08:52] hey chrisccoulson! [08:53] good morning [08:54] hey jibel! [08:54] Salut didrocks [08:59] chrisccoulson, hey, the fix to enable source from PPAs worked, I'll deploy it in the lab and will re-run ff-trunk [09:10] hey [09:11] Laney, good morning, how are you? [09:11] good - there's sun! [09:11] how are you? [09:12] good, no sun here though :-( [09:14] :( [09:15] Greetings all. [09:15] ola BigWhale [09:16] hey Laney! [09:16] Hey Laney. [09:17] seb128: move to Lyon, we have sun :) [09:17] hey BigWhale [09:17] hey dduffey [09:17] and didrocks [09:17] :)) [09:17] didrocks, sure you have sun, it's the south of France [09:17] seb128: tssssss [09:17] * didrocks opens google maps [09:17] ;-) [09:17] * Laney thinks of going there for his summer holidays ;-) [09:18] Laney: south or Lyon? :) [09:18] the Real South [09:18] 370 kms [09:18] from the pyrenees, (in longitude) [09:18] lattitude? [09:19] paris is at 240 km from the north [09:20] so paris is way more part of North than Lyon to the South [09:20] lool: take that! ^ [09:20] didrocks, we never denied that Paris is on the North, you are the one denying that Lyon is in the South :p [09:20] didrocks: I once met a Marseillais who told me that "North of Lyon, they are all Belgians" [09:21] seb128: I heard Paris is "at the center" :) [09:21] bochecha_: ahah, indeed, it's all relative :) [09:21] people in Toulouse apparently think we are still german in the East :p [09:22] :) [09:22] jibel: so so so, you know that I discovered yesterday evening that daily release was working by accident? :) [09:23] didrocks, uh? sometimes bugs fix other bugs :) what is it? [09:24] jibel: basically, when we get into the chroot, we bzr bd -S [09:24] (and we bindmount that dir) [09:25] this will create ../build-area and all other files in .. [09:25] when trying to setup my tests for those, I saw failures [09:25] like "can't write to…" and so on [09:25] and of course, those .. were not copied outside of the chroot [09:25] and we needed them [09:25] so I really wondered why it's working in prod [09:26] in fact, we also bindmount the GNUPG_DIR [09:26] and in prod (and my first tests when I wrote those), it was a parent dir of the current dir we bindmount [09:26] so everything was available :) [09:26] the current dir and its parent dir [09:27] I see [09:27] jibel: I'm writing a test to ensure we don't fall into that trap and then will fix it properly [09:28] (basically a one line change, we already pass the right directory as parameter to cd into it, we just need to bindmount the parent dir) [09:28] but a test first to ensure we fail otherwise :) [09:29] didrocks, ok, tell me if you want a review or test something [09:30] jibel: I would love to, let me finish that up and I'll ping you [09:57] Hi guys! [09:57] http://cgit.freedesktop.org/libreoffice/core/commit/?id=2b791f1cc51eaad25bd3464f94231fe4b236fae6 [09:57] \o\ [09:57] /o/ [09:58] \o/ [09:59] -j16 here I come ;-) [10:01] xnox: I laugh at your -j16, being at -j32 right now at home (http://skyfromme.wordpress.com/2012/11/12/dicke-bertha-online/) and waiting for someone with a distcc cluster to throw LibreOffice at it ... [10:01] 100k deletions? You'll never end up with a positive line of code count now.. [10:03] Sweetshark: a big "wow!", congratulations for landing this! [10:04] something like 21k is changelog/manpages/configure script. [10:06] mlankhorst: when mst came over to LibreOffice/Red Hat from Sun/Oracle (where he was sitting next to me at Sun), he introduced himself with: 'my intend it to contribute a negative linecount to LibreOffice by refactoring and cleaning up duplication'. We kinda agree there ... [10:10] it's been my goal for any project I work on too, just haven't succeeded at it :P [10:12] mlankhorst: join libreoffice, _thats_ one of the things we have plenty of opportunity for ;) (e.g. http://nabble.documentfoundation.org/Thoughts-on-killing-off-SfxDispatcher-Execute-td4039742.html ) [10:12] I do notice you talk a lot about libreoffice though! [10:14] hmmm, i think today is going to be painful. 5 minutes after starting chromium in gdb and it still hasn't finished loading the debugging symbols [10:15] 'out of swap' [10:15] heh [10:16] that's probably what will happen [10:16] it happened towards the end of linking yesterday [10:17] yeah it should be telling that the google devs had to write a linker for chrome [10:19] hmmm, i'm not sure this is actually going to work [10:19] so i wonder how i'm actually going to debug this, if i can't run it in a debugger on the target hardware? [10:23] MOAR HARDWARE [10:25] chrisccoulson: gdb over serial port? [10:25] with gdbserver [10:25] mlankhorst, would that be faster? [10:26] if it is, then i'll certainly give it a try. it can't be any worse than what i'm already doing ;) [10:26] what is your target? [10:26] mlankhorst, armhf [10:26] probably [10:27] oh, it's loaded the symbols now \o/ [10:27] only took nearly 20 minutes.... [10:27] but you need the symbols on the host machine in that case [10:28] ah, crap. gdb just aborted [10:28] * mlankhorst tries [10:29] at least I imagine you need them on the host [10:30] seems so === vrruiz_ is now known as rvr_ [10:31] i'll try gdbserver now [10:31] i can copy the 2GB chrome binary on to my local machine ;) [10:32] something like 'set sysroot /home/ccoulson/armhfchroot' would probably help too [10:33] oh lol [10:33] set sysroot remote [10:34] I fear you need a gdb that works with armhf binaries though [10:34] :( [10:35] but looks like you can build gdb with target specified in gdb/debian/target [10:41] ooh, beer delivery [10:41] it builds a nice gdb for you that installs, 'set sysroot remote:/' grabs files from the remote system it seems [10:47] start /usr/bin/arm-linux-gnueabihf-gdb 'target remote panda:port' 'set sysroot remote:/''set debug-file-directory remote:/usr/lib/debug' (I think, if you want full debug symbols) [10:58] Oh, update-notifier, I eagerly await your demise. [11:02] chrisccoulson: does it work? :-) [11:19] mlankhorst, yeah, it's working great \o/ [11:19] thanks! [11:19] np [11:20] now i can and hunt down this toolchain bug ;) [11:21] chrisccoulson, \o/ [11:24] this is so much better than trying to debug locally on the pandaboard [11:25] I wouldn't even have tried chromium locally, much courage! [11:37] chrisccoulson, did you see reports about images corruptions in firefox recently? [11:37] jibel: I have unit and system tests for running cowbuilder: https://code.launchpad.net/~didrocks/cupstream2distro/tests [11:38] jibel: the real fix is at http://bazaar.launchpad.net/~didrocks/cupstream2distro/tests/revision/229 :) [11:38] seb128, yeah, i get it too. i thought it might be a firefox regression, but if it seems that it started happening on all firefox branches at the same time (and only in raring) [11:38] didrocks, jibel: ok, so gtk 3.7 is getting close to be ready, there are a few theming issues still but that's being resolved [11:38] so it's likely a platform regression [11:38] didrocks, jibel: is there any way you could run the unity tests with the version in the ubuntu desktop ppa? [11:39] chrisccoulson, right, I was thinking the same, Rick mentioned it the other day and he's on raring, so it's not something new from the ppa [11:39] chrisccoulson, I didn't see anything obvious that changed in raring though :-( like cairo, xorg, etc didn't change [11:40] seb128, i'm not sure what it could be atm. i'd have thought cairo would be unlikely in any case, as it's not using the system cairo for drawing images [11:40] hum, right [11:41] and i turned gfx.xrender.enabled off locally as well, and it still happens [11:41] chrisccoulson, do you know when it started? [11:41] I really started noticing yesterday [11:41] but Rick mentioned it earlier in the week [11:41] seb128: no, but if you coordinate with mterry, we can copy it in the ubuntu-unity ppa and run the tests with it [11:42] didrocks, ok, will do, thanks [11:42] seb128, it's been there for quite a while. bisecting firefox has been on my todo list, but there's not much point in that if it's likely to be a platform bug [11:42] i might try a quantal install in a bit, and start selectively upgrading things [11:42] it's quite reproducible now, so it shouldn't be that hard [11:43] yw :) [11:43] didrocks, seb128 we can do it without copying to the unity-ppa I think [11:43] seb128, this is why i'd really like to get firefox tests working properly ;) [11:43] (reftests would probably catch this) [11:43] with intermittent failure [11:43] **failures [11:44] jibel: I won't really like copying it in the seed :/ [11:44] seb128, didrocks I'll copy the unity job and add the desktop ppa to the preseed in a +junk branch [11:44] jibel: if you think it's useful, but it's not scalable to doing it everytime [11:45] seb128, it's a test for a limited period right? [11:52] chrisccoulson, yeah, tests would make easy to figure at what time it started breaking [11:52] jibel, yes [11:52] jibel, I plan to upload gtk to raring this week or next week [11:52] \o/! [11:52] jibel, but I would like a round of testing to make sure it will not break unity tests first [11:53] seb128, yeah. i guess for now i'll try the selective upgrading from quantal. but after i've got chromium working first [11:53] this user jobs stuff is interesting [11:53] do we have a list of targets to convert to using them? [11:53] chrisccoulson, do you want me to try the upgrade stuff? [11:53] seb128, only if you've got time :) [11:53] Laney, we listed a few potential ones on the spec [11:53] chrisccoulson, there is always something to do, but that's true for both of us :p [11:53] seb128, what graphics hardware does rick have? [11:54] chrisccoulson, I've a vm, I can try there [11:54] chrisccoulson, we are all intel [11:54] heh [11:54] could be an intel driver issue... [11:54] ah yeah [11:54] yeah, possibly. i wonder if we can find any non-intel users with the problem :/ [11:54] did anyone on !intel noticed graphical/image corruptions in raring with firefox? [11:55] mlankhorst, ^? ;-) [11:55] no [11:55] (nvidia) [11:56] the only problem I have is that google docs makes it hang sometimes (webapps?) [11:56] nouveau? [11:56] Laney, no comment ;) [11:57] it might be useful to attach gdb to it though [11:57] anyhow I haven't noticed anything on my laptop running radeon [11:57] yeah I generally avoid trying to trigger if it possible ;-) [11:57] because then I have to kill and go find my yubikey for lastpass ... [11:58] it could be lastpass ;) [11:58] seb128, chrisccoulson: it's probably related to switching to sna acceleration, though I don't have issues and all intel here.. [11:58] :P [11:58] i think that has a reputation for being quite poor quality [11:58] suppose so [11:58] seb128: any issues in doing updates today on raring , keep getting told its only a partial upgrade [11:58] let's try [11:59] huge firefox-dbg is huge [11:59] Laney, eg, https://bugzilla.mozilla.org/show_bug.cgi?id=674535 [11:59] Mozilla bug 674535 in JavaScript Engine "Lastpass causes immortal zombie compartments" [Normal,Resolved: fixed] [12:00] slow firefox -g is slow :P [12:00] Laney, how much ram do you have? [12:01] 4G [12:01] ah, it's alright once started [12:01] valgrind works ok against xserver :D [12:02] anyone familiar with unittest here? I have a hierachy of test class inheriting ones from each other (so B inherits from A). However, if I run only B, I see that the tests for A are ran twice through nosetests [12:02] as long as you don't hit too many software fallbacks [12:02] do you think it's beacause of the import? [12:02] maybe pitti has a clue there ^ [12:03] tjaalton, thanks for the pointer btw [12:03] czajkowski, no known issues, but there is a cogl transition ... not sure if that requires to go through partial updates [12:03] shouldn't do [12:03] Laney, ^ do you know? [12:03] didrocks: notetest just walks a classes' dir() [12:03] britney should stop that [12:03] but there are some cases it doesn't catch [12:03] chrisccoulson: when did these start? [12:03] didrocks: i. e. for B it will run all tests of A as well, as it inherits A's test_* methods [12:04] tjaalton, it's been like it for some weeks now [12:04] i can't remember exactly [12:04] pitti: right, but it seems to me it's running A's one *twice* [12:04] chrisccoulson: and do you have an example? [12:04] how do you get u-m to tell you the packages involved? [12:04] czajkowski, can you "sudo apt-get upgrade" (don't validate) and pastebin the log? [12:04] sna was turned on a month ago [12:04] Laney, ^ [12:04] pitti: even if I just ask nosetests to run "B" [12:04] pitti: I would have thought, it would run B + A [12:04] but just A once, not twice [12:04] tjaalton, yeah, that seems like the sort of timeframe that it's been happening for [12:04] tjaalton, how do your turn off sna? [12:04] tjaalton, is there a way to turn that off at runtime? [12:05] what seb128 said ;) [12:05] not runtime [12:05] xorg.conf? [12:05] via xorg.conf yes [12:05] can you pastebin a file example? [12:05] I didn't edit a xorg.conf for years [12:05] I would appreciate a lazy copy/paste :p [12:05] http://pastebin.com/9AhSFLY3 [12:05] tjaalton, "example" ... screenshot or website? [12:06] seb128: something I could verify here [12:06] tjaalton, it happens randomly, I have issues on bugzilla, launchpad, etc [12:06] an url or such [12:06] hmm [12:06] screenshot would do [12:06] gah chromium crash :/ [12:06] *stab* [12:06] there is a bug open about some corruption with chromium [12:06] didrocks: yeah, that's annoying, but thats' how subclasses work; perhaps you can factorize common functinoality into an ABC and have A and B inherit from that, but not from each other? [12:06] could be the same [12:07] although that was only for an older generation of intel.. [12:07] seb128: http://pastebin.ubuntu.com/5573158/ [12:08] czajkowski, what about "dist-upgrade"? [12:08] pitti: hum, sorry, I think I didn't explain well [12:08] pitti: basically, I want that B runs A tests [12:08] pitti: but what happens is that B is running A tests twice [12:08] tjaalton, http://ubuntuone.com/61fpONo2YyaP6tL52kyhJp [12:08] when I'm asking nose to run B tests [12:08] that's the comment box on bugzilla [12:08] didrocks: oh, I see; no idea about that, I'm afraid [12:09] seb128: http://pastebin.ubuntu.com/5573162/ [12:09] chrisccoulson: http://people.canonical.com/~laney/firefox.log is a thread apply all bt while hung that's interesting to you [12:09] pitti: ok, that's really really weird, B is just "A" with a different setUp() [12:09] I think my extensions are lastpass/pentadactyl/adblock plus [12:09] chrisccoulson: huh, not seen that [12:09] chrisccoulson: but if you have a bug open, you can toss it over to xserver-xorg-video-intel I guess [12:09] czajkowski: those 2 removals are intentional, they've been replaced by newer versions [12:10] ok [12:10] hmm [12:10] u-m shouldn't give a partial upgrade for that [12:10] didn't it get some fixes in that regard recently? [12:10] upstream is actively closing/fixing the intel bugs on lp, which is pretty neat [12:11] czajkowski: can you check this with cjwatson please? He was looking at the previous partial upgrade bug [12:11] tjaalton, i'll try without sna first before reporting an intel bug [12:11] chrisccoulson: sure, thanks [12:11] in #ubuntu-devel would be good [12:11] czajkowski: first check you have update-manager 1:0.183 [12:12] Laney, hmmm, that trace looks quite normal. what does "call DumpJSStack()" give? [12:13] pitti: I think I got it, as I was importing BaseUnitTest which is in __init__.py of the module, it imported as well all the tests from the package and rerun them [12:13] basically the import screw nosetests :/ [12:14] aah [12:17] something very wrong when you go to tab complte cjwatson and notice he's not here [12:17] :( [12:18] bah, unity fail [12:18] screenshot made both my screen go white, I had to restart the session [12:18] tjaalton, chrisccoulson: http://ubuntuone.com/13BzkJ2r7ARDsEQQU7mr9O example if corruption [12:20] seb128, yeah, that's pretty much what i see too [12:21] seb128: what hw is this? [12:21] tjaalton, i5 [12:21] the cpu/gpu [12:22] lspci|grep VGA [12:22] Intel(R) Core(TM) i5 CPU M 520 @ 2.40GHz [12:22] 00:02.0 VGA compatible controller: Intel Corporation Core Processor Integrated Graphics Controller (rev 02) [12:22] so ironlake (gen5) [12:23] don't see the bug with gen4/6/7 :) [12:24] chrisccoulson: check yours ^ [12:24] 1 second (although, i suspect it's identical) ;) [12:24] tjaalton, [12:24] févr. 25 18:30:46 but I get the firefox issues on my other computer also, which is i945 [12:24] tjaalton, just as a piece of info [12:25] that's gen3 [12:25] i'm lucky then :) [12:25] Intel(R) Core(TM) i5 CPU M 560 @ 2.67GHz [12:25] 00:02.0 VGA compatible controller: Intel Corporation Core Processor Integrated Graphics Controller (rev 02) [12:25] lol [12:25] heh [12:25] i've got a faster CPU :P [12:25] anyway, check with uxa, and report the bug with ubuntu-bug xorg [12:25] if doesn't happen with it [12:26] or directly u-b x-x-v-intel [12:26] should dtrt [12:32] didrocks, who can i ping for questions concerning ubuntu tablet and phone [12:34] hum, timeouted [12:34] seif: depends on what the questions is :) [12:34] seif: ogra knows about the deployement, as sergio [12:36] ogra_, so how many changes happened since the official release of ubuntu tablet [12:36] seif, ? [12:37] seif, for the android side, read the git logs, for the ubuntu rootfs read the package changelogs ... [12:38] (its the same as in any ubuntu ... plus the android git) [12:39] ogra_, link to the ubuntu change logs [12:39] /usr/share//chanelog...... === seb128_ is now known as seb128 [12:42] seif, beyond that https://launchpad.net/ubuntu-touch-preview has the bzr trees (btw, this discuaaion should wither be on the tablet ML or in #ubuntu-tablet) [12:42] *either [12:46] ogra_: seif #ubuntu-touch [12:46] tablet and phone channels were merged into one [12:46] err, sorry, yes i know [12:46] seif just confused me with his questions :) [12:47] -tablet is a redirect anyway though [13:07] tjaalton, sorry I lost track with my IRC timeouting, is there any extra info you want me to provide or anything I should test? [13:07] seb128: just that if you can reproduce the corruption with that xorg.conf [13:08] if not, file a bug with 'ubuntu-bug xorg' === soren__ is now known as soren [13:11] tjaalton, ok, I've that in place, let's see [13:11] chrisccoulson, going to test as well? [13:17] seb128, i can do. although i don't want to kill my current session just yet :) [13:17] Ran 83 tests in 449.521s [13:17] \o/ [13:17] (this long time is because I ran the system tests as well) [13:17] units only: Ran 57 tests in 0.671s [13:17] and fixed a regression + pushed :) [13:18] nice [13:18] chrisccoulson: I don't think call DumpJSStack() is returning :( [13:18] Laney, oh, that's not good [13:18] ok, I really should play with upstart session, /me opt in for that ppa [13:19] don't forget to choose the right session to actually use it ;-) [13:19] * Laney forgot that at first [13:20] Laney, good point, thanks ;-) [13:20] I'm doing a bunch of environment dumps first [13:20] so I can diff and look for differences [13:20] looking at whether it can be made to work with all gnome-sessions [13:20] I guess you need to conditionalise some of the jobs on $DESKTOP_SESSION [13:21] or look in the .session file or something === smspilla1 is now known as smspillaz [13:28] hi hackers [13:28] desrt, hey, how are you? [13:28] good [13:28] how's my package? :) [13:28] * desrt tries to remember if he selected the "Overnight Delivery" option... [13:29] hm the more kernels I build the less I care about extra options in the config, and just accept the defaults [13:29] desrt, you didn't :p on my todo but I didn't get to it yet ;-) [13:29] fair enough :) [13:29] need to restart my session with new upstart and xorg uxa [13:29] brb === Ursinha-afk is now known as Ursinha [13:57] seb128, the ppa from which you want to test packages is ppa:ubuntu-desktop/ppa? [13:58] jibel, correct [13:58] * desrt is off the PPAs [13:58] that PPA man... one harsh drug [13:58] messed me up bad [13:58] thats bad juju that stuff [13:58] seb128, good, I started a first run to make sure everything is ok. [13:58] (pun not intended) [13:58] jibel, excellent, thanks [13:59] smspillaz: give any thought to all that work i tried to pile on you recently? :) [13:59] desrt: it turned out I had this 5000 word paper due in a week or two [13:59] desrt, the desktop ppa has gtk 3.7.10 and jibel is nice enough to make a run of unity tests with that version to see if there is any issue [13:59] seb128: i'd normally be excited by that, but i have a serious case of PPA fever :p [13:59] ;-) [13:59] desrt: also it is quite close to FF and I probably want to wait until some of my other changes go in first [14:00] desrt, if the testing goes fine and thar you poke larsu enough to get the theming problems solved it might land in raring this week [14:00] desrt: which would include much increased rendering performance on nvidia, which we would have been "first" to was the review process not so slow [14:00] seb128: it may interest you to know that larsu hates css now :) [14:00] haha [14:01] (mutter got support for buffer_age not long ago because someone just went ahead and did it, and there was no three month long review backup to wait for) [14:02] well [14:02] obviously that's because the mutter hackers are so brilliant that they don't need reviews [14:02] desrt: yep [14:02] the scales are clearly tipped in their favor *sigh* [14:03] i hate when someone replies to sarcasm with what may or may not be further sarcasm [14:03] * desrt is always left wondering what to do next [14:03] desrt: I'm Jay Leno [14:04] (as it is, usually the answer is to complain awkwardly...) [14:04]